Building a company? Whatever works

I don't think it matters much whether your company is remote or not

So I’ve been running Remote for about seven years right now, and before this I spent five years at GitLab. Both companies stand out because they are run in very different ways compared to standard practices. GitLab already started out as a fully remote, fully distributed team and we did everything in public. We had a public handbook, which is still available at handbook.gitlab.com.

Everything had to be documentation first. If you wanted to change company policy, you would make a merge request and change the company handbook before announcing it. This was a pretty effective way to discuss making particular changes. For example, you could suggest a change in the text and then have a discussion on the merge request before it was incorporated into the handbook. Once it was in the handbook, it became official—that was the new policy.

GitLab ran in a very peculiar way. It was very different from other companies. When we started Remote we took significant inspiration from how GitLab was run. Remote, of course, is also a fully remote and distributed company. We also have a public handbook, which is available at remote.com/handbook.

But we also made quite a few changes in the way we ran the company. Some of it was based on the fact that it’s simply a different company with a different product. Some of it was based on me being a different person than Sid (CEO GitLab at the time), and wanting to do things differently.

All of this to say that these are companies that are both objectively successful, but are run in a very different way from how organizations are typically run.

We’re very used to companies running from an office. They tend to have very specific and well-known processes and ways that they are structured. We now have sufficient evidence to believe that there are vastly different ways of running companies that could still produce very successful businesses.

Beyond my personal experience, I believe you can observe similar things in how very large (public) companies are run. I’m sure you’ve heard about Jensen Huang from NVIDIA, who doesn’t do one-on-ones and has many, many reports. Then there are companies like Revolut being run in an extremely data-driven way, and Meta before that - where data is god.
Inversely, yesterday I spoke with a leader at a company that focuses very much on people above all else, giving them ownership and freedom to set and achieve goals. This company too is wildly succesful.

This leads me to believe that there’s not really a best practice for running a company. I think most of the “best practices” are simply the retelling of the story of a successful company. I bet that the commonality between successful companies is the fact that the leaders are intentional about the way they run the company, iterating and changing it over time.
Successful companies seem to succeed not by a particular method, but rather by strong conviction above everything else.

this is just a pretty picture I took the other day

It doesn’t really matter that you’re remote

This is a long-winded way of me saying that I don’t actually think remote-based companies or office-based companies are fundamentally better off that the other. I think it depends largely on what suits the founders and/or the leadership of a company. If that team is intentional about how they create the company they are building, they will be successful whether they are in office, whether they work remotely, or whether they follow particular methods or invent something new altogether.

I do think there are particular things that companies should or shouldn’t do explicitly. For example, I think a company should be able to make quick decisions and should be able to move fast. Things that stop a company from doing that are likely hurting the company. It’s going to be real hard to create and maintain a successful business if that isn’t true. So all of the above is assuming the company does everything else really, really well.

Companies of the future

If you zoom out further, there must be a company that functions in a way that is further from what companies like Remote and GitLab have been doing.
Some companies are experimenting with 4-day work-weeks, and others trying to be much more flexible around employment (although I’d suggest you avoid misclassifying people as contractors).

The most interesting example around pay is certainly Gumroad by Sahil Lavignia: he has designed a scheme that allows paid-by-hour contractors to decide whether to get paid in all cash or cash and a chosen amount of equity over which Gumroad pays dividends. He wrote a really great blog post on this that I recommmend reading.

Are companies going to be actually distributed and fully merit based? I think some yes - but not most. It’s still fun to work together people, and there certainly is value in working with the same group of people for a long time. As someone that runs a company I’d also not be very excited to not know whether I still have a workforce from one day to the next.

That is until everything is just robots.